CN110431564A - Method and system for Fingerprint enhancement - Google Patents
Method and system for Fingerprint enhancement Download PDFInfo
- Publication number
- CN110431564A CN110431564A CN201880004911.8A CN201880004911A CN110431564A CN 110431564 A CN110431564 A CN 110431564A CN 201880004911 A CN201880004911 A CN 201880004911A CN 110431564 A CN110431564 A CN 110431564A
- Authority
- CN
- China
- Prior art keywords
- fingerprint image
- image data
- fingerprint
- filter
- filtering
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 34
- 238000001914 filtration Methods 0.000 claims abstract description 59
- 230000002708 enhancing effect Effects 0.000 claims abstract description 51
- 238000013480 data collection Methods 0.000 claims abstract description 48
- 230000008859 change Effects 0.000 claims description 13
- 238000004590 computer program Methods 0.000 claims description 6
- 230000006870 function Effects 0.000 description 24
- 239000011159 matrix material Substances 0.000 description 10
- 238000012804 iterative process Methods 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 238000003860 storage Methods 0.000 description 4
- 239000003990 capacitor Substances 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 238000001514 detection method Methods 0.000 description 2
- 230000005684 electric field Effects 0.000 description 2
- 230000005611 electricity Effects 0.000 description 2
- 230000014759 maintenance of location Effects 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 238000013475 authorization Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000003989 dielectric material Substances 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 125000000524 functional group Chemical group 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000015654 memory Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 238000002604 ultrasonography Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/20—Image enhancement or restoration using local operators
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/253—Fusion techniques of extracted features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/80—Fusion, i.e. combining data from various sources at the sensor level, preprocessing level, feature extraction level or classification level
- G06V10/806—Fusion, i.e. combining data from various sources at the sensor level, preprocessing level, feature extraction level or classification level of extracted features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/12—Fingerprints or palmprints
- G06V40/13—Sensors therefor
- G06V40/1306—Sensors therefor non-optical, e.g. ultrasonic or capacitive sensing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/12—Fingerprints or palmprints
- G06V40/1347—Preprocessing; Feature extraction
- G06V40/1359—Extracting features related to ridge properties; Determining the fingerprint type, e.g. whorl or loop
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/50—Maintenance of biometric data or enrolment thereof
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Multimedia (AREA)
- Human Computer Interaction (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- General Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Computing Systems (AREA)
- Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- General Engineering & Computer Science (AREA)
- Image Input (AREA)
- Collating Specific Patterns (AREA)
- Image Processing (AREA)
Abstract
The present invention relates to a kind of methods for Fingerprint enhancement, including by the first low-pass filter (302；And the first weight (306 402)；402 ') application (S704) is in original fingerprint image data (300), to generate the first filtering fingerprint image data collection (310).By the second low-pass filter (304；And the second weight (308 403)；403 ') application (S706) is in original fingerprint image data, to generate the second filtering fingerprint image data collection (312).The filter coefficient of second filter is different from the filter coefficient of first filter.First filtering fingerprint image data collection and the second filtering fingerprint image data collection (S708) are combined, to generate final enhancing fingerprint image.Electronic device (100) the invention further relates to fingerprint sensing system and including fingerprint sensing system.
Description
Technical field
The present invention relates to the methods for Fingerprint enhancement.The invention further relates to fingerprint sensing system and including fingerprint
The electronic device of sensing system.
Background technique
Various types of biological recognition systems are used increasingly, to provide the safety and/or enhancing that enhance
Convenience for users.
Particularly, fingerprint sensing system uses in such as consumer electronics device, this is attributed to the fact that their small shape
Pattern, high-performance and customer acceptance.
Fingerprint sensor generally includes picture element matrix, which is configured to sense the fingerprint pattern of finger.It collects
The signal in each of the pixel element is simultaneously followed by subsequent processing to form fingerprint image.It is desirable that final fingerprint image is low noise
Sound pitch resolution ratio fingerprint image, the fingerprint image can be used for fingerprint recognition and apply and can relatively quickly be obtained.
However, being usually distorted in different ways by the original fingerprint image that fingerprint sensor captures.For example, original fingerprint
Image may be fuzzy, non-uniform, or comprising the noise from each provenance, such as the electricity of the simulation from fingerprint sensor
Gain stage in road.
Fuzzy in fingerprint image may be as caused by the electric field change between finger and fingerprint sensor region.US
9,235,283 disclose a kind of example of finger biometric identification sensor comprising deblurring circuit is to mitigate in fingerprint image
Fuzzy distortion.In general, executing deblurring by the influence for attempting reversion changing electric field using different pattern functions.One
The disadvantage is that deblurring may cause the artifact in the other parts of image.
However, fuzzy one of several distortions of quality for only influencing fingerprint image, therefore still can be further
Improve fingerprint image quality.
Summary of the invention
In view of the above and other disadvantage of the prior art, the object of the present invention is to provide a kind of for Fingerprint enhancement
Method has the fingerprint image for improving quality to provide.
According to the first aspect of the invention, a kind of method for Fingerprint enhancement is provided, this method comprises: obtaining
The original fingerprint image data captured by fingerprint sensor arrangement；First low-pass filter and the first weight are applied to original finger
For print image data to generate the first filtering fingerprint image data collection, the first low-pass filter includes at least one set of filter coefficient,
By the second low-pass filter and the second weight applied to original fingerprint image data to generate the second filtering fingerprint image data collection,
Second low-pass filter includes at least one set of filter coefficient, and wherein this group of filter factor of second filter is different from the first filter
Fingerprint image data collection after first filtering fingerprint image data collection and the second filtering is carried out group by this group of filter coefficient of wave device
It closes to generate final enhancing fingerprint image.
The present invention is based on following implemented: at least two low-pass filters and two weights are respectively applied to original fingerprint figure
As data, to obtain at least two different filtering fingerprint image data collection.To filtering fingerprint image data collection be combined with
Generate the fingerprint image of enhancing.It can handle the inhomogeneities in such as fingerprint image in this way, and utilize improved property
It can be that fingerprint recognition application program generates clearer image.
Therefore, compared with art methods, the present invention can specifically provide the fingerprint image of enhancing, has and reduces
Inhomogeneities, and may have the acutance of less noise and raising.Therefore, improved fingerprint recognition performance is possible
's.
The filter coefficient that the filter coefficient of second filter is different from first filter can be each filter
At least one filter coefficient between coefficient sets is different.
" obtaining original fingerprint image data " includes obtaining the original fingerprint image data previously captured.However, it may be used also
To include capture original fingerprint image data and generate final enhancing fingerprint image with it.
Original fingerprint image data is the data captured by fingerprint sensor, and is preferably without experience any influence figure
The processing of image quality amount.Original fingerprint image data may be constructed such that the Two-Dimensional Moment of the data point of the pixel from fingerprint sensor
Battle array.Data point from pixel can be the intensity value in gained fingerprint image.
By first filtering fingerprint image data collection and second filtering fingerprint image data collection be combined may include with by
Each data set is added together by point mode.For example, if filtering data collection is configured to the matrix of filtering data, it can be to square
Battle array summation is to generate the fingerprint image data for finally enhancing fingerprint image.In other words, the first filtering of combination fingerprint image
Data set and the second filtering fingerprint image data can be the individual element summation to the intensity value from fingerprint image data collection.
Low-pass filter can be the spatial low-pass filter across 2 dimension fingerprint image data applications.If tieing up filter 1
In the case where be shown as figure, then low-pass filter can be by describing the function representation of " curve ", or if ties up filters 2
In the case where be shown as figure, then low-pass filter can be by describing the function representation on " surface ".The shape on curve or surface by
Filter coefficient limits, to define the characteristic of low-pass filter.
Furthermore, it is possible to provide the discrete representation of low-pass filter in the form of including the matrix of filter coefficient.
In some embodiments, filter is Gaussian filter.Gaussian filter has the response of Gaussian function form
Function.Gaussian filter itself is known to technical staff.About the above content, the filter coefficient of Gaussian filter can be by
Gaussian function export.2 dimension Gaussian functions define the shape on " surface " in the case of 2 dimensions, and in the case of 1 dimension Gaussian function definition, 1 dimension
The shape of " curve ".
In the case where application Gaussian filter, the filter coefficient of each filter is defined by Gaussian function, and can
With according to standard deviation by single state modulator.A standard deviation is selected for first filter, is second filter choosing
Select another standard deviation.Therefore, the filter coefficient of two filters is different.
Embodiment according to the present invention, the first weight can be with the second weighteds.It in other words, can be using difference
Weight come generate filtering fingerprint image data collection.By properly selecting weight, which advantageously provide final enhancing fingerprints
The quality of image further improved, for example, clearer image, the image with less noise and/or final enhancing
The uniformity of the raising of fingerprint image.
Weight can be positive number or negative, and be multiplied advantageous by by weight with corresponding filtering fingerprint image data
To apply.
In one embodiment, multiple low-pass filters and weight can be applied to original fingerprint image data, with
Multiple filtering fingerprint image data collection are generated, wherein multiple filtering finger print data collection are combined, are referred to generating final enhancing
Print image.Therefore, original fingerprint image data set can pass through the combination (bank) of low-pass filter and weight.Using additional
Low-pass filter and weight, it might even be possible to further increase the quality of final enhancing fingerprint image.
Embodiment according to the present invention, method may include being repeated by successive ignition by low-pass filter and weight
Applied to the final enhancing fingerprint image generated in corresponding previous ones, the final enhancing fingerprint image of update is iteratively generated
Picture.Therefore, finally enhancing fingerprint image is fed back to low-pass filter, and is used as the input of low-pass filter as " former
Begin " fingerprint image data, iteratively to generate the final enhancing fingerprint image of update.Iteration is recyclable to be had several times to generate
The final enhancing fingerprint image of the picture quality further improved.
When generating final enhancing fingerprint image using iterative process, groups of filter coefficients can have change between iterations
Change.For example, the first group of filter coefficient used in the first iteration can change as the third used in successive iterations
Group filter coefficient.Similarly, it is the 4th group of filter coefficient that second group of filter coefficient can change between iteration twice.
In addition, at least one of weight can have between iterations when iteration is to generate final enhancing fingerprint image
Variation.
Filter coefficient can be selected based on the attribute of the original fingerprint image data of acquirement.Carrying out such selection is
In order to, such as improve the acutance of final enhancing fingerprint image or improve bio-identification performance, in order in finger print identifying program
Fingerprint recognition performance is improved as authentication image using final enhancing fingerprint image.
The selection of filter coefficient can be carried out based on the empirically determined selection from offline machine learning, this is offline
Machine learning is related with the final enhancing fingerprint image obtained according to the selection of original fingerprint image data and filter coefficient.It should
It selects to be also based on using original fingerprint image data as input and be referred to using finally enhancing as the expectation of target
Print image solves the problems, such as mathematical optimization.
In the mode similar with the selection of above-mentioned filter coefficient, it is also based on the original fingerprint image data obtained
Attribute select weight.For example, can choose weight to obtain the final enhancing fingerprint image with the acutance improved.
According to the second aspect of the invention, a kind of fingerprint sensing system is provided, which includes: fingerprint sensor dress
It sets, is used to obtain fingerprint image data；And control unit, it is configured to execute any in any above embodiment
A step.
Fingerprint sensor can be capacitive fingerprint sensing device.Such capacitive fingerprint sensing device may include condenser type
Sensing element, each capacitance type sensing element provide indicate the specific sensing element and touch sensor surface finger surface it
Between capacity coupled measurement.Sensing element at position corresponding with the ridge in fingerprint by than with the paddy phase in fingerprint
Sensing element at corresponding position shows stronger and finger capacitive coupling.
However, various embodiments of the invention are not limited to the finger of the sensing element including the use of particular fingerprint detection technology
Line sensing device, but be equally applicable to such as optics, calorifics, capacitor, acoustics/ultrasound or piezoelectricity fingerprint sensor.
Fingerprint sensor arrangement may include the fingerprint pattern for being provided by sensing element instruction sensing signal into
The adjunct circuit of row operation.Such adjunct circuit can be for example including sample circuit and analog to digital conversion circuit.Therefore, fingerprint passes
Sensor is so as to providing the fingerprint pattern signal as digital signal.Alternatively, fingerprint pattern signal can be used as analog signal
It is provided.For example, signal can be the analog or digital value of instructed voltage, the signal so that can with by finger, (or finger is examined
Other conductive bodies near geodesic structure), the capacitor of capacitor that constitutes of finger detection structure and dielectric material between them
It is proportional.
The fingerprint pattern of sensing can be used for various purposes, such as bio-identification registration or certification, or be based on fingerprint pattern
Navigation etc..
Further embodiment of the invention and the effect obtained by the second aspect of the invention with largely
It is upper be similar to above for described in the first aspect of the present invention those.
A kind of electronic device is further provided, which includes: fingerprint sensing system；Wherein control unit is matched
It is set to: the certification request for being directed to finger is provided to fingerprint sensing system；Checking signal is received from fingerprint sensing systems；And if
Authentication signal instruction authenticates successfully, then executes at least one movement.
Electronic device can be mobile device (such as mobile phone), smart card, tablet computer or laptop computer,
Or desktop computer or any other existing or future the suitable electronic device including fingerprint sensor.
A kind of computer program product is further provided, which includes: computer-readable medium, is had on it
The computer program for being used to control electronic device of storage, which includes: fingerprint sensor, is configured to obtain original
Beginning fingerprint image data and control unit are configured to receive the original fingerprint image data captured by fingerprint sensor,
Wherein computer program product include: for by the first low-pass filter and the first weight be applied to original fingerprint image data with
The code of the first filtering fingerprint image data collection is generated, the first low-pass filter includes: at least one set of filter coefficient, and being used for will
Second low-pass filter and the second weight are applied to original fingerprint image data to generate the second filtering fingerprint image data collection
Code, the second low-pass filter include at least one set of filter coefficient, and wherein the filter coefficient of second filter is different from
The filter coefficient of first filter, and for the first filtering fingerprint image data collection and second to be filtered fingerprint image number
It is combined according to collection to generate the code of final enhancing fingerprint image.
In short, this method includes by the first low-pass filter the present invention relates to a kind of method for Fingerprint enhancement
It is applied to original fingerprint image data with the first weight to generate the first filtering fingerprint image data collection.By the second low-pass filter
It is applied to original fingerprint image data with the second weight to generate the second filtering fingerprint image data collection.The filtering of second filter
Device coefficient is different from the filter coefficient of first filter.By the first filtering fingerprint image data collection and the second filtering fingerprint image
Data set is combined, to generate final enhancing fingerprint image.The invention further relates to fingerprint sensing system and including fingerprint sense
The electronic device of examining system.
When studying accessory claim and being described below, it will be understood that further characteristic of the invention and excellent
Point.Those skilled in the art recognizes, can be to different characteristic of the invention without departing substantially from the scope of the present invention
It is combined to generate the implementation embodiment other than embodiments described below.
Detailed description of the invention
Now with reference to the attached drawing for showing example embodiment of the invention, the present invention will be described in more detail these and its
Its aspect, in which:
Fig. 1 schematically shows the electronic device of example embodiment according to the present invention；
Fig. 2 schematically shows the fingerprint acquisition apparatus in Fig. 1；
Fig. 3 conceptually illustrates the embodiments of the present invention for generating final enhancing fingerprint image；
Fig. 4 conceptually illustrates the embodiments of the present invention for generating final enhancing fingerprint image；
Fig. 5 conceptually illustrates the embodiment party of the invention for iteratively generating final enhancing fingerprint image
Formula；
Fig. 6 a conceptually illustrates low-pass filter function；
Fig. 6 b conceptually illustrates one group of filter coefficient；
Fig. 7 is the flow chart for schematically showing the method for embodiment according to the present invention；
Fig. 8 is the flow chart for schematically showing the method for embodiment according to the present invention；And
Fig. 9 is the flow chart for schematically showing the method for embodiment according to the present invention.
Specific embodiment
In this detailed description, described according to the present invention referring especially to having the mobile device of composition fingerprint acquisition apparatus
Fingerprint sensing system and method various embodiments.It is to be noted, however, that the electronic device of many other types can
Such fingerprint acquisition apparatus, such as tablet computer, desktop computer, laptop computer, smart card etc. can be integrated with.
Turning now to attached drawing and particularly, Fig. 1, which is schematically shown, to be configured to answer in the form of mobile device 100
With the example according to the electronic device of the concept of present disclosure, which has 102 He of fingerprint sensor of composition
Display unit 104 with touch screen interface 106.In this embodiment, fingerprint sensor 102 is disposed in mobile device
100 front side, display unit 104 also are located at the front side.Fingerprint sensor 102 for example can be used to unlock mobile device 100 with
And/or person is used to authorize the transaction etc. executed using mobile device 100.Fingerprint sensor 102 can certainly be placed on shifting
The back side or side of dynamic device 100.
Preferably and as apparent to those skilled in the art, mobile device 100 shown in Fig. 1 further comprises
It is single for the first antenna of WLAN/Wi-Fi communication, the second antenna for telecommunication, microphone, loudspeaker and phone control
Member.Certainly, mobile device may include further hardware element.
It shall yet further be noted that present invention may apply to be related to the electronic device of any other type, such as laptop computer,
Remote controler, tablet computer, the smart card including fingerprint sensor or any other type it is existing or future similarly configured
Device, including any kind of IoT (Internet of Things) device, wherein it is desirable to the user for allowing to be realized specifically be arranged and/
Or identification/certification of user.
About all electronic devices, such as electronic device shown in Fig. 1, fingerprint sensor 102 is arranged to and fits
Control unit for controlling fingerprint sensor 102 is electrical or wirelessly communicates.Control unit may include microprocessor, microcontroller
Device, programmable digital signal processor or other programming devices.Control unit can with or alternatively include dedicated forming electricity
Road, programmable gate array or programmable logic array, programmable logic device or digital signal processor.Include in control unit
Programming device, such as the place of above-mentioned microprocessor, microcontroller or programmable digital signal processor, processor can be into
One step includes controlling the computer-executable code of the operation of programmable device.It should be understood that by means of control unit (or usually
As " processing circuit " discuss) provide function all or some part can at least partly with fingerprint sensor 102 collect
At.
Therefore, fingerprint acquisition apparatus 102 can be included in the fingerprint sensing system including control unit (not shown).
Control unit may be configured to execute the step of describing referring to Fig. 3-8 and function.Control unit can be independent fingerprint sensing
A part or control unit of system can be the control unit of main electronic device, i.e. Setup Controller.Such device control
Device processed can be further configured to provide the certification request for being directed to finger to fingerprint acquisition apparatus 102.Such certification request
It can for example be related to requesting access to the i.e. electronic device or any other device using fingerprint sensor, or authorization transaction etc..
Control unit then can receive authentication signal from fingerprint sensor arrangement 102.If authentication signal instruction authenticates successfully, can
To execute at least one movement.
Referring to Fig. 2, the view of fingerprint sensor 102 slightly amplified is conceptually illustrated.Using capacitance type sensing
In the case where technology, fingerprint sensor 102 is configured to include a large amount of sensing element, is preferably disposed in two-dimensional array.
Two-dimensional array can have the size that the plan of depending on is realized, and in embodiments, use 160 × 160 pixels.Other rulers
It is very little being certainly possible to and within the scope of the invention, including compared with above-mentioned example with the two-dimensional array of less pixel.
Single sense element (also being indicated as pixel) is indicated in Fig. 2 by appended drawing reference 202.
Fig. 3 conceptually illustrates the implementation of the invention for generating enhancing fingerprint image from original fingerprint image 300
Mode.By the original fingerprint image 300 that fingerprint acquisition apparatus captures be input to low-pass filter 302,304 and weight 306,
308 combination (bank) 301, original fingerprint image data 300 is by the first low-pass filter 302 and the first weight 306, to produce
Raw first filtering fingerprint image data collection 310.First low-pass filter 302 includes first group of the characteristic of description filter 302
Filter coefficient.
Original fingerprint image data 300 also passes through the second low-pass filter 304.In addition, the second weight 308 is applied to original
Beginning fingerprint image data 300, to generate the second filtering fingerprint image data collection 312.First filtering 310 He of fingerprint image data collection
Second filtering fingerprint image data collection 312 generates parallel, i.e., original fingerprint image data 300 is input to two individually parallel
In filter.
Second low-pass filter 304 includes second group of filter coefficient, is different from first group of filter coefficient.In other words
It says, original fingerprint image data is independently input in two different filters, this two different filters generate not
Same output filters fingerprint image data collection.In addition, in some embodiments, weight 306 and 308 is also different from each other.For example,
First weight 306 can be " -1 ", and the second weight 308 can be " 1 ".Weight and corresponding filtering fingerprint image data collection phase
Multiply.In general, weight can be any positive number or negative.
Then, the first filtering fingerprint image data collection 310 and the second filtering fingerprint image data collection 312 are combined,
To generate final enhancing fingerprint image 314.It may include to filtering fingerprint that filtering fingerprint image data 310,312, which is combined,
Image data 310,312 is summed.Image data usually indicates (for example, forming each pixel) by the matrix of image intensity value, and
And the summation of image data is the direct operation to Matrix Calculating sum.
Fig. 4 conceptually illustrates further embodiment of the invention, and wherein filter bank (bank) is shown as
Including multiple low-pass filters and weight.Original fingerprint image data 300 is input to low-pass filter (402-N) and weight
Each of (402 '-N ').Therefore, for each of low-pass filter and weight group, filtering fingerprint image number is generated
According to collection.In the conceptual representation, two filtering fingerprint image data collection 310 and 312 are illustrated only, but as mentioned, it deposits
In the filtering fingerprint image data collection generated from each pair of low-pass filter (402-N) and weight (402 '-N ').For example, such as
Simultaneously therefore there are X weight (402 '-N ') there are X low-pass filter (402-N) for fruit, then therefore there are X filtering fingerprint images
As data set, these data sets are then combined to form final enhancing fingerprint image 314.
Fig. 5 conceptually illustrates another example embodiment.The embodiment of Fig. 5 includes low-pass filter (402-
) and weight shown in Fig. 4 N.In addition, Fig. 5 includes the diagram for generating the iterative process of final enhancing fingerprint image 314.Therefore,
Primary or iteration, the final enhancing generated after being combined to multiple filtering fingerprint image data collection more times is referred to
Print image gives the filter bank 301 including low-pass filter 402-N and weight 402 '-N ' as input feedback.Therefore, finally
Enhance fingerprint image 314 by low-pass filter 402-N and weight 402'-N' to generate one group of new filtering fingerprint image data
These data sets are combined to form the final enhancing fingerprint image updated by collection.Iterative process can be repeated as many times, this takes
Certainly in the characteristic of such as original fingerprint image data.
In addition, filter coefficient and weight group can change between iterations in iterative process shown in Fig. 5.In
In some embodiments, only one group in each filter and weight coefficient group can change between iterations.For example, weight
402 ' (or any other weight) can change between iteration twice.Similarly, low-pass filter 402 (or any other
Low-pass filter) groups of filter coefficients can be changed between iteration twice.
Change filter coefficient by selecting one group of new filter coefficient from the groups of filter coefficients of multiple storages.
Filter is used in software-based platform, this is usually using such as Gaussian filter or other low-pass filters to figure
The case where being filtered as data.
Low-pass filter of the invention can be various filter types, but in a kind of possible realization, low pass filtered
Wave device is Gaussian filter.Gaussian filter its be usually characterized by the response letter of the Gaussian function form with standard deviation
Number.Using Gaussian filter, standard deviation is the filter characteristic that can be selected, and then passes through Gaussian function
Number defines groups of filter coefficients.
Therefore, Gaussian function defines one group of filter coefficient.Gaussian filter itself be those skilled in the art
Know, and Gaussian function (referring also to Fig. 6 b) can mathematically be indicated in two dimensions with following formula:
Wherein σ is the standard deviation of Gaussian function.In order to be filtered to fingerprint image, by the Gauss with standard deviation
Function is applied to fingerprint image.Gaussian filtering may be implemented as the 2 dimension cores with the image convolution to be filtered.
Another possibility is in the both direction across fingerprint image using 1 dimension filter.For example, can refer to first
Using 1 dimension filter in the horizontal direction of print image, it is then followed by the vertical direction of image using 1 dimension filter.Certainly
1D filter can also be applied in vertical direction first, then apply 1D filter in the horizontal direction.
Fig. 6 a conceptually illustrates low-pass filter function 602, and Fig. 6 b shows the matrix of filter coefficient 606
One group of filter coefficient (only one is numbered) of 604 forms.Low-pass filter function 602 may, for example, be height as described above
This function.
Filter coefficient 606 is conceptually illustrated in Fig. 6 b.It can be exported from low-pass filter function 602, Huo Zheke
Filter coefficient is individually chosen with picture quality based on expectations output.The quantity of filter coefficient can change.For example, only
It gives some instances, matrix 604 may include the filter coefficient of 5x5,7x7,9x9,12x12 or 15x15.Matrix 604 is not necessarily
It must be square matrix.For filtering operation, core, the i.e. matrix 604 of filter coefficient 606 and image data carry out convolution.
Fig. 7 shows the flow chart of the method and step of embodiment according to the present invention, and the invention discloses one kind to be used for
The method of Fingerprint enhancement.In first step S701, the original fingerprint picture number captured by fingerprint sensor arrangement is obtained
According to.
In step S704, the first low-pass filter and the first weight are applied to original fingerprint image data, to generate
First filtering fingerprint image data collection, the first low-pass filter include at least one set of filter coefficient.For example, in low-pass filter
It is to select in the embodiment of Gaussian filter filter coefficient, to provide the Gaussian Profile for indicating Gaussian filter
Desired standard deviation.
In step S706, the second low-pass filter and the second weight are applied to original fingerprint image data, to generate
Second filtering fingerprint image data collection.Second low-pass filter includes at least one set of filter coefficient, wherein second filter
Groups of filter coefficients is different from the groups of filter coefficients of first filter.
Then, in step S708, by first filtering fingerprint image data collection and second filtering fingerprint image data collection into
Row combination, to generate final enhancing fingerprint image.
Optionally, referring now to Fig. 8, including step S702, to select to be applied to the weight of original fingerprint image data.
In addition, including optional step S703, to select the filter coefficient of low-pass filter (such as Gaussian filter).
The selection of filter coefficient and/or weight can be based on analyzing original fingerprint image data.For example,
It can be selected to obtain the final enhancing fingerprint image sharpened, or obtain final enhancing fingerprint image more evenly, or
The combination of enhancing.Another method is to optimize biological knowledge based on the tradeoff between optimization erroneous matching rate and wrong probability of mismatch
Other performance.
In the presence of for determining the various process for selecting which weight and/or filter coefficient.Understand and how to carry out this
A kind of mode of kind selection is to execute offline machine learning from the original fingerprint image data previously captured, how will be former with study
Beginning fingerprint image data is changed to desired final enhancing fingerprint image.Therefore, by being analyzed simultaneously a large amount of prior images
It attempts to enhance image by adjusting filter coefficient and/or weight, can identify how the original that will be inputted with training system
Beginning fingerprint image data is changed to enhancing fingerprint image.Such study can for example using have supervision or unsupervised machine
Device learning algorithm is realized.
Fig. 9 is the flow chart of the method and step of another embodiment according to the present invention.The step of Fig. 7 and 8, is also included in
In embodiment in Fig. 9, and it is not described further referring again to Fig. 9.
The schematic illustration of the iterative process for generating final enhancing fingerprint image is presented in Fig. 9.Therefore, when
When generating final enhancing fingerprint image in step S708, as input feedback to filter, i.e., finally enhance fingerprint image
By filter step S704 and S706, and select in the step S702 and S703 in new filter coefficient and each iteration
The optional step of new weight.If finally enhancing fingerprint image after step S708 is satisfactorily, to end processing simultaneously
The final enhancing fingerprint image of output, the final enhancing fingerprint image otherwise updated be also used as input feedback to step S704 and
S706, wherein with optional step S702 and S703.
Process in Fig. 7-9 shows the illustrative embodiments with two weights and two filters.Such as Fig. 4 and
Shown in 5, the quantity of weight and filter can be any quantity, and be certainly not limited to two.For example, filter and weight
Quantity can be 2,3,4,5,6,7,8,9,10 etc..
The function of present disclosure can use existing computer processor to realize, or by being used for appropriate system
Special purpose computer processor realize, be incorporated to for this or another purpose, or realized by hard-wired system.
It include program product scope of the present disclosure interior embodiment, which includes being stored in it for being loaded with or having
On machine-executable instruction or data structure machine readable media.Such machine readable media can be can be by general
Or any usable medium that special purpose computer or other machines with processor access.For example, this machine readable Jie
Matter may include RAM, ROM, EPROM, EEPROM, CD-ROM or other disc memories, magnetic disk storage or other magnetic storages
Device, or can be used for being loaded with or storing any other medium of desired program code, form be machine-executable instruction or
Data structure, and can be accessed by general or specialized computer or other machines with processor.When by network or its
He communicates to connect (hardwired, wireless or hardwired or wireless combination) to machine transmission or when providing information, and machine is correctly
Connection is considered as machine readable media.Therefore, any this connection is appropriately referred to as machine readable media.Combinations of the above
It is included in the range of machine readable media.Machine-executable instruction include for example make general purpose computer, special purpose computer or
Dedicated processes machine executes the instruction and data of specific function or functional group.
Although attached drawing can show sequence, the sequence of step can be different from the sequence of description.Can also simultaneously or
Part is performed simultaneously two or more steps.This variation will depend on selected software and hardware system and depend on
Designer's selection.All these variations are all scope of the present disclosure interior.Likewise it is possible to using standard programming technology come complete
At software realization, which has rule-based logic and other logics, to complete various Connection Steps, processing step, ratio
Compared with step and steps in decision-making.In addition, even if describing the present invention referring to the specific illustrative embodiment of the present invention, still
For a person skilled in the art, many different changes, modifications etc. will become obvious.
In addition, those skilled in the art is practicing by the research to attached drawing, disclosure and appended claims
When claimed invention it will be appreciated that with the change to the embodiment of disclosure is realized.In addition, in the claims, word
Language " comprising " is not excluded for other elements or step, and indefinite article "one" be not excluded for it is multiple.
Claims (16)
1. a kind of method for Fingerprint enhancement, which comprises
Obtain the original fingerprint image data (300) that (S701) is captured by fingerprint sensor arrangement；
By the first two-dimensional low-pass filter (302；And the first weight (306 402)；402 ') application (S704) is in the original finger
Print image data, to generate the first filtering fingerprint image data collection (310), first two-dimensional low-pass filter includes at least one
Group filter coefficient,
By the second two-dimensional low-pass filter (304；And the second weight (308 403)；403 ') application (S706) is in the original finger
Print image data, to generate the second filtering fingerprint image data collection (312), second two-dimensional low-pass filter includes at least one
Group filter coefficient, wherein the group that this group of filter coefficient of the second filter is different from the first filter filters
Device coefficient,
The first filtering fingerprint image data collection and the second filtering fingerprint image data collection are combined (S708),
To generate final enhancing fingerprint image (314).
2. according to the method described in claim 1, wherein first weighted is in second weight.
3. the method according to any one of the preceding claims, comprising:
Multiple two-dimensional low-pass filters (402-N) and multiple weights (402-N ') are applied to the original fingerprint image data,
To generate multiple filtering fingerprint image data collection, wherein the multiple filtering fingerprint image data collection is combined, to generate
The final enhancing fingerprint image.
4. the method according to any one of the preceding claims, comprising:
It repeats two-dimensional low-pass filter and weight being applied to the institute generated in corresponding previous ones by successive ignition
Final enhancing fingerprint image is stated, the final enhancing fingerprint image of update is iteratively generated.
5. according to the method described in claim 4, wherein at least one set in each group filter coefficient has between iteration twice
Variation.
6. method according to claim 4 or 5, wherein at least one of described weight has change between iteration twice
Change.
7. the method according to any one of the preceding claims, comprising:
Based on the characteristic of acquired fingerprint image data, (S703) described two-dimensional low-pass filter coefficient is selected.
8. the method according to any one of the preceding claims, comprising:
Based on acquired fingerprint image data, (S702) is selected to be used for the weight of the filter.
9. having described in the acutance improved most according to the method described in claim 8, wherein selecting the weight for obtaining
Enhancing fingerprint image eventually.
10. the method according to any one of the preceding claims, wherein each of described two-dimensional low-pass filter is
Corresponding Gaussian filter.
11. being consequently for according to the method described in claim 10, wherein the standard deviation of the Gaussian filter is different from each other
The filter coefficient of the Gaussian filter is different from each other.
12. a kind of fingerprint sensing system, comprising:
Fingerprint sensor arrangement (102), is used to obtain fingerprint image data；And
Control unit is configured to execute the step of any one of preceding claims.
13. fingerprint sensing system according to claim 12, wherein the fingerprint sensor arrangement is that capacitance type fingerprint passes
Sensor.
14. a kind of electronic device (100), comprising:
Fingerprint sensing system described in any one of 2 or 13 according to claim 1；Wherein described control unit is configured to:
- Xiang Suoshu fingerprint acquisition apparatus provides the certification request for being directed to finger；
Authentication signal is received from the fingerprint acquisition apparatus；And
If the authentication signal instruction authenticates successfully, at least one movement is executed.
15. electronic device according to claim 14, wherein the electronic device is mobile device (100) or smart card.
16. a kind of computer program product, including computer-readable medium, the computer-readable medium has in the calculating
What is stored on machine readable medium is used to control the computer program of electronic device, and the electronic device includes fingerprint sensor, institute
It states fingerprint sensor and is configured to obtain original fingerprint image data and control unit, described control unit is configured to connect
The original fingerprint image data captured by the fingerprint sensor is received, wherein the computer program product includes:
It is used for the first two-dimensional low-pass filter (302；And the first weight (306 402)；402 ') it is applied to the original fingerprint
For image data to generate the code of the first filtering fingerprint image data collection (310), first two-dimensional low-pass filter includes extremely
Few one group of filter coefficient；
It is used for the second two-dimensional low-pass filter (304；And the second weight (308 403)；403 ') it is applied to the original fingerprint
For image data to generate the code of the second filtering fingerprint image data collection (312), second two-dimensional low-pass filter includes extremely
Few one group of filter coefficient, wherein this group of filter coefficient of the second filter is different from the group of the first filter
Filter coefficient, and
For the first filtering fingerprint image data collection and the second filtering fingerprint image data collection to be combined
(S708) to generate the final code for enhancing fingerprint image (314).
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| SE1751271 | 2017-10-13 | ||
| SE1751271-6 | 2017-10-13 | ||
| PCT/SE2018/051008 WO2019074423A1 (en) | 2017-10-13 | 2018-10-02 | Method and system for fingerprint image enhancement |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110431564A true CN110431564A (en) | 2019-11-08 |
| CN110431564B CN110431564B (en) | 2020-11-17 |
Family
ID=66101014
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201880004911.8A Active CN110431564B (en) | 2017-10-13 | 2018-10-02 | Method and system for fingerprint image enhancement |
Country Status (4)
| Country | Link |
|---|---|
| US (1) | US11080546B2 (en) |
| EP (1) | EP3695342B1 (en) |
| CN (1) | CN110431564B (en) |
| WO (1) | WO2019074423A1 (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10706254B2 (en) * | 2016-03-22 | 2020-07-07 | Shenzhen GOODIX Technology Co., Ltd. | Method and apparatus for calibrating fingerprint image, and terminal |
| CN112329681A (en) * | 2020-11-13 | 2021-02-05 | 北京思比科微电子技术股份有限公司 | Filtering method applied to fingerprint identification |
Families Citing this family (25)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10445547B2 (en) | 2016-05-04 | 2019-10-15 | Invensense, Inc. | Device mountable packaging of ultrasonic transducers |
| US10315222B2 (en) | 2016-05-04 | 2019-06-11 | Invensense, Inc. | Two-dimensional array of CMOS control elements |
| US10441975B2 (en) | 2016-05-10 | 2019-10-15 | Invensense, Inc. | Supplemental sensor modes and systems for ultrasonic transducers |
| US10562070B2 (en) | 2016-05-10 | 2020-02-18 | Invensense, Inc. | Receive operation of an ultrasonic sensor |
| US10706835B2 (en) | 2016-05-10 | 2020-07-07 | Invensense, Inc. | Transmit beamforming of a two-dimensional array of ultrasonic transducers |
| US10539539B2 (en) | 2016-05-10 | 2020-01-21 | Invensense, Inc. | Operation of an ultrasonic sensor |
| US11673165B2 (en) | 2016-05-10 | 2023-06-13 | Invensense, Inc. | Ultrasonic transducer operable in a surface acoustic wave (SAW) mode |
| US10452887B2 (en) | 2016-05-10 | 2019-10-22 | Invensense, Inc. | Operating a fingerprint sensor comprised of ultrasonic transducers |
| US10984209B2 (en) | 2017-12-01 | 2021-04-20 | Invensense, Inc. | Darkfield modeling |
| WO2019104719A1 (en) * | 2017-12-01 | 2019-06-06 | 深圳市汇顶科技股份有限公司 | Fingerprint image enhancement method and fingerprint image module |
| US10997388B2 (en) | 2017-12-01 | 2021-05-04 | Invensense, Inc. | Darkfield contamination detection |
| US11151355B2 (en) | 2018-01-24 | 2021-10-19 | Invensense, Inc. | Generation of an estimated fingerprint |
| US10755067B2 (en) | 2018-03-22 | 2020-08-25 | Invensense, Inc. | Operating a fingerprint sensor comprised of ultrasonic transducers |
| WO2020263875A1 (en) | 2019-06-24 | 2020-12-30 | Invensense, Inc. | Fake finger detection using ridge features |
| US11216632B2 (en) * | 2019-07-17 | 2022-01-04 | Invensense, Inc. | Ultrasonic fingerprint sensor with a contact layer of non-uniform thickness |
| US11176345B2 (en) * | 2019-07-17 | 2021-11-16 | Invensense, Inc. | Ultrasonic fingerprint sensor with a contact layer of non-uniform thickness |
| US11232549B2 (en) | 2019-08-23 | 2022-01-25 | Invensense, Inc. | Adapting a quality threshold for a fingerprint image |
| US11392789B2 (en) | 2019-10-21 | 2022-07-19 | Invensense, Inc. | Fingerprint authentication using a synthetic enrollment image |
| EP4100176B1 (en) | 2020-03-09 | 2024-10-09 | InvenSense, Inc. | Ultrasonic fingerprint sensor with a contact layer of non-uniform thickness |
| US11922719B2 (en) * | 2020-03-30 | 2024-03-05 | Nec Corporation | Image processing apparatus, image processing method, and recording medium |
| US11328165B2 (en) | 2020-04-24 | 2022-05-10 | Invensense, Inc. | Pressure-based activation of fingerprint spoof detection |
| CN113918908A (en) * | 2020-07-07 | 2022-01-11 | 三星电子株式会社 | Method and apparatus for fingerprint verification |
| US11995909B2 (en) | 2020-07-17 | 2024-05-28 | Tdk Corporation | Multipath reflection correction |
| CN112800996A (en) * | 2020-09-09 | 2021-05-14 | 神盾股份有限公司 | Electronic device with fingerprint sensing function and fingerprint comparison method |
| US11810391B2 (en) * | 2021-05-06 | 2023-11-07 | Novatek Microelectronics Corp. | Method for image processing circuit and related sampling circuit |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5602934A (en) * | 1993-09-08 | 1997-02-11 | The Regents Of The University Of California | Adaptive digital image signal filtering |
| US20020146178A1 (en) * | 2000-09-01 | 2002-10-10 | International Business Machines Corporation | System and method for fingerprint image enchancement using partitioned least-squared filters |
| TW530485B (en) * | 2000-09-08 | 2003-05-01 | Sanyo Electric Co | Horizontal contour signal generating circuit for single board type color camera |
| US20030112374A1 (en) * | 2003-01-31 | 2003-06-19 | Samsung Electronics Co., Ltd. | Method and apparatus for image detail enhancement using filter bank |
| WO2003079274A1 (en) * | 2002-03-20 | 2003-09-25 | Philips Intellectual Property & Standards Gmbh | Method of improving fingerprint images |
| CN101276464A (en) * | 2007-03-26 | 2008-10-01 | 株式会社东芝 | Picture processing apparatus and method |
| CN101329726A (en) * | 2008-07-30 | 2008-12-24 | 电子科技大学 | Method for reinforcing fingerprint image based on one-dimensional filtering |
| EP2026586A1 (en) * | 2006-05-24 | 2009-02-18 | Panasonic Corporation | Image processing device |
| CN102103692A (en) * | 2011-03-17 | 2011-06-22 | 电子科技大学 | Fingerprint image enhancing method |
| CN103824060A (en) * | 2014-02-28 | 2014-05-28 | 清华大学 | Method for extracting fingerprint detail points |
| CN104715459A (en) * | 2015-03-27 | 2015-06-17 | 浙江大学 | Blood vessel image enhancement method |
| US20160307023A1 (en) * | 2015-04-15 | 2016-10-20 | Samsung Electronics Co., Ltd. | Method and apparatus for recognizing fingerprint |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104298962B (en) | 2014-09-04 | 2018-04-27 | 深圳市汇顶科技股份有限公司 | Finger print detection device and method |
| US9641820B2 (en) * | 2015-09-04 | 2017-05-02 | Apple Inc. | Advanced multi-band noise reduction |
| TWI591545B (en) * | 2016-06-30 | 2017-07-11 | Fingerprint image detecting device and method thereof |
-
2018
- 2018-10-02 CN CN201880004911.8A patent/CN110431564B/en active Active
- 2018-10-02 US US16/753,762 patent/US11080546B2/en active Active
- 2018-10-02 WO PCT/SE2018/051008 patent/WO2019074423A1/en unknown
- 2018-10-02 EP EP18865664.9A patent/EP3695342B1/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5602934A (en) * | 1993-09-08 | 1997-02-11 | The Regents Of The University Of California | Adaptive digital image signal filtering |
| US20020146178A1 (en) * | 2000-09-01 | 2002-10-10 | International Business Machines Corporation | System and method for fingerprint image enchancement using partitioned least-squared filters |
| TW530485B (en) * | 2000-09-08 | 2003-05-01 | Sanyo Electric Co | Horizontal contour signal generating circuit for single board type color camera |
| WO2003079274A1 (en) * | 2002-03-20 | 2003-09-25 | Philips Intellectual Property & Standards Gmbh | Method of improving fingerprint images |
| US20030112374A1 (en) * | 2003-01-31 | 2003-06-19 | Samsung Electronics Co., Ltd. | Method and apparatus for image detail enhancement using filter bank |
| EP2026586A1 (en) * | 2006-05-24 | 2009-02-18 | Panasonic Corporation | Image processing device |
| CN101276464A (en) * | 2007-03-26 | 2008-10-01 | 株式会社东芝 | Picture processing apparatus and method |
| CN101329726A (en) * | 2008-07-30 | 2008-12-24 | 电子科技大学 | Method for reinforcing fingerprint image based on one-dimensional filtering |
| CN102103692A (en) * | 2011-03-17 | 2011-06-22 | 电子科技大学 | Fingerprint image enhancing method |
| CN103824060A (en) * | 2014-02-28 | 2014-05-28 | 清华大学 | Method for extracting fingerprint detail points |
| CN104715459A (en) * | 2015-03-27 | 2015-06-17 | 浙江大学 | Blood vessel image enhancement method |
| US20160307023A1 (en) * | 2015-04-15 | 2016-10-20 | Samsung Electronics Co., Ltd. | Method and apparatus for recognizing fingerprint |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10706254B2 (en) * | 2016-03-22 | 2020-07-07 | Shenzhen GOODIX Technology Co., Ltd. | Method and apparatus for calibrating fingerprint image, and terminal |
| CN112329681A (en) * | 2020-11-13 | 2021-02-05 | 北京思比科微电子技术股份有限公司 | Filtering method applied to fingerprint identification |
Also Published As
| Publication number | Publication date |
|---|---|
| EP3695342A1 (en) | 2020-08-19 |
| WO2019074423A1 (en) | 2019-04-18 |
| US20200285882A1 (en) | 2020-09-10 |
| EP3695342A4 (en) | 2020-12-02 |
| CN110431564B (en) | 2020-11-17 |
| EP3695342B1 (en) | 2024-09-25 |
| US11080546B2 (en) | 2021-08-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110431564A (en) | Method and system for Fingerprint enhancement | |
| US11508146B2 (en) | Convolutional neural network processing method and apparatus | |
| EP3333768A1 (en) | Method and apparatus for detecting target | |
| EP3065085B1 (en) | Digital image processing using convolutional neural networks | |
| TWI596551B (en) | Method and electronic device for noise mitigation | |
| Liu et al. | Texture classification using spectral histograms | |
| Laparra et al. | Iterative gaussianization: from ICA to random rotations | |
| CN110235139A (en) | Method for authenticating the finger of the user of electronic device | |
| CN106030613A (en) | Method and fingerprint sensing system for authenticating a candidate fingerprint | |
| CN113570508A (en) | Image restoration method and device, storage medium and terminal | |
| CN110765857A (en) | Fingerprint identification method, chip and electronic device | |
| CN107251052A (en) | Method and fingerprint sensing system for forming fingerprint image | |
| Ren et al. | Enhanced non-local total variation model and multi-directional feature prediction prior for single image super resolution | |
| CN111784665A (en) | OCT image quality assessment method, system and device based on Fourier transform | |
| CN110796250A (en) | Convolution processing method and system applied to convolutional neural network and related components | |
| CN106503743A (en) | A kind of quantity is more and the point self-adapted clustering method of the high image local feature of dimension | |
| Feng et al. | BM3D‐GT&AD: an improved BM3D denoising algorithm based on Gaussian threshold and angular distance | |
| CN109784233B (en) | Phase-locked value weighted spatial filtering method and device based on Lp-norm | |
| Lyu et al. | Identifiability-guaranteed simplex-structured post-nonlinear mixture learning via autoencoder | |
| Fermanian et al. | PnP-ReG: Learned regularizing gradient for plug-and-play gradient descent | |
| JP6526494B2 (en) | INFORMATION PROCESSING PROGRAM AND INFORMATION PROCESSING DEVICE | |
| Yu et al. | Bionic RSTN invariant feature extraction method for image recognition and its application | |
| JP6366420B2 (en) | Data creation program, data creation method, and data creation device | |
| US11495049B2 (en) | Biometric feature reconstruction method, storage medium and neural network | |
| CN109727262A (en) | The system and method for stream for processes data values |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20211202 Address after: Gothenburg Patentee after: Fingerprint kaana Kadun Intellectual Property Co.,Ltd. Address before: Gothenburg Patentee before: FINGERPRINT CARDS AB |
|
| TR01 | Transfer of patent right |